www.nature.com/scientificreports
OPEN
received: 07 July 2015 accepted: 09 November 2015 Published: 22 December 2015
Body part-centered and full bodycentered peripersonal space representations Andrea Serino1,2, Jean-Paul Noel1,2,†, Giulia Galli1,2,3, Elisa Canzoneri1,2, Patrick Marmaroli4, Hervé Lissek4 & Olaf Blanke1,2,5 Dedicated neural systems represent the space surrounding the body, termed Peripersonal space (PPS), by integrating visual or auditory stimuli occurring near the body with somatosensory information. As a behavioral proxy to PPS, we measured participants’ reaction time to tactile stimulation while task-irrelevant auditory or visual stimuli were presented at different distances from their body. In 7 experiments we delineated the critical distance at which auditory or visual stimuli boosted tactile processing on the hand, face, and trunk as a proxy of the PPS extension. Three main findings were obtained. First, the size of PPS varied according to the stimulated body part, being progressively bigger for the hand, then face, and largest for the trunk. Second, while approaching stimuli always modulated tactile processing in a space-dependent manner, receding stimuli did so only for the hand. Finally, the extension of PPS around the hand and the face varied according to their relative positioning and stimuli congruency, whereas the trunk PPS was constant. These results suggest that at least three body-part specific PPS representations exist, differing in extension and directional tuning. These distinct PPS representations, however, are not fully independent from each other, but referenced to the common reference frame of the trunk. Neurophysiological studies in monkeys described a pool of multisensory neurons-mainly in ventral premotor cortex, ventral intraparietal area, area 7, and putamen – dedicated to represent the space surrounding the body, termed Peripersonal space (PPS)1. Typically these neurons have a tactile receptive field (RF) centered on a specific body part (hand, arm, face, trunk or shoulder, with some neurons having RFs covering even the entire body) and a visual and/or an auditory RF that overlaps spatially with the tactile RF and extends in depth for a distance ranging from 5 to 100 cm from the body2–4. Neuropsychological studies in patients5 and behavioral studies in healthy participants6 provided first evidence for the existence of a PPS system in the human brain. These studies showed that tactile perception is more strongly modulated by visual or auditory stimuli7,8 when these are presented close, as compared to far, from the body. Neuroimaging studies9,10 associated these effects with neuronal processing in fronto-parietal brain regions, homologues to the regions hosting PPS neurons in non-human primates. Whereas, the properties of PPS neurons in monkeys in terms of receptive field size, response preferences, and spatial frames of reference have been extensively described, knowledge about the characteristics of PPS representation in humans is mostly limited to the evidence of stronger multisensory interaction near the hand or the face. Therefore, it is still debated to what degree PPS representation in the human brain is based on the same properties as those implemented in monkey PPS neurons11,12. The aim of the present study is to measure behavioral responses in humans reflecting the concept and the properties of RFs in primate PPS neurons. To this aim, we applied a recently developed paradigm in which participants are requested to respond as fast as possible to tactile stimulation administered on a part of their body, while 1
Laboratory of Cognitive Neuroscience, Brain Mind Institute, School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland. 2Center for Neuroprosthetics, School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland. 3Department of Psychology, Sapienza University, Rome, Italy. 4 Laboratory of Electromagnectics and Acoustics, Institute of Electrical Engineering, School of Engineering, Ecole Polytechnique Federale de Lausanne, Lausanne, Switzerland. 5Department of Neurology, University Hospital, Geneva, Switzerland. †Present Address: Vanderbilt Brain Institute, Vanderbilt University Medical Center, Vanderbilt University, Nashville, TN, USA. Correspondence and requests for materials should be addressed to A.S. (email: [email protected]) Scientific Reports | 5:18603 | DOI: 10.1038/srep18603
1
www.nature.com/scientificreports/ task-irrelevant approaching or receding external cues (auditory or visual stimuli) are presented13–16. On each trial, the tactile stimulus is administered at a different temporal delay from the cue onset and thus touch is processed when the cues are perceived at a different distance from the participant’s body. In this way, we determine the critical distance at which an external stimulus affects tactile processing. This point can be considered as the boundary of PPS representation in humans. In a series of 7 experiments (164 participants), we applied this method to study the properties of the human PPS. By administering tactile stimuli to different body parts, we firstly compared the extension of PPS representation around the hand, face, and trunk. Then, we studied the relationships between discrete PPS representations around different body parts, by either varying the relative position of the arm and the trunk, while sounds approached the hand or the trunk (Experiments 1–5), or by varying the congruency between tactile stimulation on the face and trunk, while auditory or visual stimuli either congruently approached the stimulated body part or incongruently the other body part (Experiments 6–7). Taken together, our findings suggest that at least three body-part specific PPS representations exist (differing in extension and directional tuning), which are all referenced to the more common reference frame of the trunk.
Results
Section 1. The relationship between Peri-trunk and Peri-hand PPS. Experiment 1. Peri-trunk PPS. The first experiment aimed at investigating PPS representation around the trunk. We tested whether an auditory stimulus interacted more strongly with a tactile stimulation on the chest when presented within a specific spatial range from the trunk. We determined the farthest point in space in which a sound significantly speeded up tactile RT as compared to a baseline unimodal tactile RT, and this point was taken as a proxy for the boundary of peri-trunk PPS. To this aim, participants were requested to respond as fast as possible to a tactile target administered to their trunk (chest, at sternum level), while task-irrelevant sounds either approached or receded from their trunk. On each trial, the tactile target was presented at one out of six possible delays from sound onset, i.e. when sounds were perceived at one out of six possible distances (D1, D2, D3, D4, D5, and D6) between 5 and 100 cm from the body (D1 through D6 respectively corresponding to 5, 24, 43, 62, 81, and 100 cm). Baseline trials (in which no auditory stimuli was provided) and catch trials (in which no tactile stimulation was given) were also acquired (see Fig. 1; see also Methods section for further details). Averaged tactile RTs for each of the six possible distances were corrected for RT to unimodal tactile stimuli. We firstly searched for a difference in multisensory RT at consecutive distances, to show a spatially dependent modulation of tactile processing and then we compared corrected RT at each distance to baseline (by definition, zero) in order to identify the farthest point at which a multisensory facilitation occurred as the boundary of peri-trunk representation (all subsequent studies follow the same logic). Baseline-corrected RTs were submitted to a 2 (Sound Direction: Looming vs. Receding) × 6 (Sound Distance: D1 through D6) Within-Subjects ANOVA. As illustrated in Fig. 2A, findings showed a significant Sound Direction × Sound Distance interaction (F(5, 65) = 18.743, p
OPEN
received: 07 July 2015 accepted: 09 November 2015 Published: 22 December 2015
Body part-centered and full bodycentered peripersonal space representations Andrea Serino1,2, Jean-Paul Noel1,2,†, Giulia Galli1,2,3, Elisa Canzoneri1,2, Patrick Marmaroli4, Hervé Lissek4 & Olaf Blanke1,2,5 Dedicated neural systems represent the space surrounding the body, termed Peripersonal space (PPS), by integrating visual or auditory stimuli occurring near the body with somatosensory information. As a behavioral proxy to PPS, we measured participants’ reaction time to tactile stimulation while task-irrelevant auditory or visual stimuli were presented at different distances from their body. In 7 experiments we delineated the critical distance at which auditory or visual stimuli boosted tactile processing on the hand, face, and trunk as a proxy of the PPS extension. Three main findings were obtained. First, the size of PPS varied according to the stimulated body part, being progressively bigger for the hand, then face, and largest for the trunk. Second, while approaching stimuli always modulated tactile processing in a space-dependent manner, receding stimuli did so only for the hand. Finally, the extension of PPS around the hand and the face varied according to their relative positioning and stimuli congruency, whereas the trunk PPS was constant. These results suggest that at least three body-part specific PPS representations exist, differing in extension and directional tuning. These distinct PPS representations, however, are not fully independent from each other, but referenced to the common reference frame of the trunk. Neurophysiological studies in monkeys described a pool of multisensory neurons-mainly in ventral premotor cortex, ventral intraparietal area, area 7, and putamen – dedicated to represent the space surrounding the body, termed Peripersonal space (PPS)1. Typically these neurons have a tactile receptive field (RF) centered on a specific body part (hand, arm, face, trunk or shoulder, with some neurons having RFs covering even the entire body) and a visual and/or an auditory RF that overlaps spatially with the tactile RF and extends in depth for a distance ranging from 5 to 100 cm from the body2–4. Neuropsychological studies in patients5 and behavioral studies in healthy participants6 provided first evidence for the existence of a PPS system in the human brain. These studies showed that tactile perception is more strongly modulated by visual or auditory stimuli7,8 when these are presented close, as compared to far, from the body. Neuroimaging studies9,10 associated these effects with neuronal processing in fronto-parietal brain regions, homologues to the regions hosting PPS neurons in non-human primates. Whereas, the properties of PPS neurons in monkeys in terms of receptive field size, response preferences, and spatial frames of reference have been extensively described, knowledge about the characteristics of PPS representation in humans is mostly limited to the evidence of stronger multisensory interaction near the hand or the face. Therefore, it is still debated to what degree PPS representation in the human brain is based on the same properties as those implemented in monkey PPS neurons11,12. The aim of the present study is to measure behavioral responses in humans reflecting the concept and the properties of RFs in primate PPS neurons. To this aim, we applied a recently developed paradigm in which participants are requested to respond as fast as possible to tactile stimulation administered on a part of their body, while 1
Laboratory of Cognitive Neuroscience, Brain Mind Institute, School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland. 2Center for Neuroprosthetics, School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland. 3Department of Psychology, Sapienza University, Rome, Italy. 4 Laboratory of Electromagnectics and Acoustics, Institute of Electrical Engineering, School of Engineering, Ecole Polytechnique Federale de Lausanne, Lausanne, Switzerland. 5Department of Neurology, University Hospital, Geneva, Switzerland. †Present Address: Vanderbilt Brain Institute, Vanderbilt University Medical Center, Vanderbilt University, Nashville, TN, USA. Correspondence and requests for materials should be addressed to A.S. (email: [email protected]) Scientific Reports | 5:18603 | DOI: 10.1038/srep18603
1
www.nature.com/scientificreports/ task-irrelevant approaching or receding external cues (auditory or visual stimuli) are presented13–16. On each trial, the tactile stimulus is administered at a different temporal delay from the cue onset and thus touch is processed when the cues are perceived at a different distance from the participant’s body. In this way, we determine the critical distance at which an external stimulus affects tactile processing. This point can be considered as the boundary of PPS representation in humans. In a series of 7 experiments (164 participants), we applied this method to study the properties of the human PPS. By administering tactile stimuli to different body parts, we firstly compared the extension of PPS representation around the hand, face, and trunk. Then, we studied the relationships between discrete PPS representations around different body parts, by either varying the relative position of the arm and the trunk, while sounds approached the hand or the trunk (Experiments 1–5), or by varying the congruency between tactile stimulation on the face and trunk, while auditory or visual stimuli either congruently approached the stimulated body part or incongruently the other body part (Experiments 6–7). Taken together, our findings suggest that at least three body-part specific PPS representations exist (differing in extension and directional tuning), which are all referenced to the more common reference frame of the trunk.
Results
Section 1. The relationship between Peri-trunk and Peri-hand PPS. Experiment 1. Peri-trunk PPS. The first experiment aimed at investigating PPS representation around the trunk. We tested whether an auditory stimulus interacted more strongly with a tactile stimulation on the chest when presented within a specific spatial range from the trunk. We determined the farthest point in space in which a sound significantly speeded up tactile RT as compared to a baseline unimodal tactile RT, and this point was taken as a proxy for the boundary of peri-trunk PPS. To this aim, participants were requested to respond as fast as possible to a tactile target administered to their trunk (chest, at sternum level), while task-irrelevant sounds either approached or receded from their trunk. On each trial, the tactile target was presented at one out of six possible delays from sound onset, i.e. when sounds were perceived at one out of six possible distances (D1, D2, D3, D4, D5, and D6) between 5 and 100 cm from the body (D1 through D6 respectively corresponding to 5, 24, 43, 62, 81, and 100 cm). Baseline trials (in which no auditory stimuli was provided) and catch trials (in which no tactile stimulation was given) were also acquired (see Fig. 1; see also Methods section for further details). Averaged tactile RTs for each of the six possible distances were corrected for RT to unimodal tactile stimuli. We firstly searched for a difference in multisensory RT at consecutive distances, to show a spatially dependent modulation of tactile processing and then we compared corrected RT at each distance to baseline (by definition, zero) in order to identify the farthest point at which a multisensory facilitation occurred as the boundary of peri-trunk representation (all subsequent studies follow the same logic). Baseline-corrected RTs were submitted to a 2 (Sound Direction: Looming vs. Receding) × 6 (Sound Distance: D1 through D6) Within-Subjects ANOVA. As illustrated in Fig. 2A, findings showed a significant Sound Direction × Sound Distance interaction (F(5, 65) = 18.743, p
